Organizations often employ large fleets of computing devices for various reasons, such as adding functionality, scalability, and availability. As the systems become larger and more complex, monitoring these systems to ensure proper operation also becomes more complex. An organization's operations, for instance, may be supported by a data center or even multiple data centers that include computer systems, networking devices, and other systems. Further, the architecture of distributed systems can involve multiple layers of dependency, often causing operational issues with one system that may affect numerous other systems. A single point of failure, for instance, can cause multiple other systems to fail or otherwise have degraded performance. Various techniques have been employed to monitor the systems. While such techniques facilitate the identification of problems and their solutions, as the number of devices being monitored grows, efficiently identifying the sources of problems and their solutions becomes more difficult.